tips<-reshape2::tips # Load dataset on tipping behavior included with reshape2 package
attributes(tips) # Check attributes of the tips dataset (names, row.names, class)
## $names
## [1] "total_bill" "tip" "sex" "smoker" "day"
## [6] "time" "size"
##
## $row.names
## [1] "1" "2" "3" "4" "5" "6" "7" "8" "9" "10" "11"
## [12] "12" "13" "14" "15" "16" "17" "18" "19" "20" "21" "22"
## [23] "23" "24" "25" "26" "27" "28" "29" "30" "31" "32" "33"
## [34] "34" "35" "36" "37" "38" "39" "40" "41" "42" "43" "44"
## [45] "45" "46" "47" "48" "49" "50" "51" "52" "53" "54" "55"
## [56] "56" "57" "58" "59" "60" "61" "62" "63" "64" "65" "66"
## [67] "67" "68" "69" "70" "71" "72" "73" "74" "75" "76" "77"
## [78] "78" "79" "80" "81" "82" "83" "84" "85" "86" "87" "88"
## [89] "89" "90" "91" "92" "93" "94" "95" "96" "97" "98" "99"
## [100] "100" "101" "102" "103" "104" "105" "106" "107" "108" "109" "110"
## [111] "111" "112" "113" "114" "115" "116" "117" "118" "119" "120" "121"
## [122] "122" "123" "124" "125" "126" "127" "128" "129" "130" "131" "132"
## [133] "133" "134" "135" "136" "137" "138" "139" "140" "141" "142" "143"
## [144] "144" "145" "146" "147" "148" "149" "150" "151" "152" "153" "154"
## [155] "155" "156" "157" "158" "159" "160" "161" "162" "163" "164" "165"
## [166] "166" "167" "168" "169" "170" "171" "172" "173" "174" "175" "176"
## [177] "177" "178" "179" "180" "181" "182" "183" "184" "185" "186" "187"
## [188] "188" "189" "190" "191" "192" "193" "194" "195" "196" "197" "198"
## [199] "199" "200" "201" "202" "203" "204" "205" "206" "207" "208" "209"
## [210] "210" "211" "212" "213" "214" "215" "216" "217" "218" "219" "220"
## [221] "221" "222" "223" "224" "225" "226" "227" "228" "229" "230" "231"
## [232] "232" "233" "234" "235" "236" "237" "238" "239" "240" "241" "242"
## [243] "243" "244"
##
## $class
## [1] "data.frame"
# Create an object of class "lm" (linear model), regressing tip on some covariates
tips.reg<-lm(formula=tip~total_bill+sex+smoker+day+time+size, data=tips)
attributes(tips.reg) # Check attributes of the tips.reg object (names, class)
## $names
## [1] "coefficients" "residuals" "effects" "rank"
## [5] "fitted.values" "assign" "qr" "df.residual"
## [9] "contrasts" "xlevels" "call" "terms"
## [13] "model"
##
## $class
## [1] "lm"
plot(tips) # Calls plotting method for class of tips dataset ("data.frame")
plot(tips.reg, which=1:2) # Calls plotting method for class of tips.reg objects ("lm"), print first two plots only
xyplot(tips) # Attempt in lattice to automatically plot objects of class "data.frame"
## Error: no applicable method for 'xyplot' applied to an object of class
## "data.frame"
ggplot(data=tips)+geom_point() # Attempt in ggplot to automatically plot objects of class "data.frame"
## Error: 'where' is missing
xyplot(tips.reg) # Attempt in lattice to automatically plot objects of class "lm"
## Error: no applicable method for 'xyplot' applied to an object of class
## "lm"
ggplot(data=tips.reg)+geom_point() # Attempt in ggplot to automatically plot objects of class "lm"
## Error: 'where' is missing
For more info: http://www.ipw.unibe.ch/content/team/klaus_armingeon/comparative_political_data_sets/index_eng.html
a) faster (though only noticeable over many and large plots)
b) simpler (at first)
c) better at trellis graphs
d) able to do 3d graphs
a) generally more elegant
b) more syntactically logical (and therefore simpler, once you learn it)
c) better at grouping
d) able to interface with maps
The general call for lattice graphics looks something like this:
graph_type(formula, data=, [options])
The specifics of the formula differ for each graph type, but the general format is straightforward
y # Show the distribution of y
y~x # Show the relationship between x and y
y~x|A # Show the relationship between x and y conditional on the values of A
y~x|A*B # Show the relationship between x and y conditional on the combinations of A and B
z~y*x # Show the 3D relationship between x, y, and z
The general call for ggplot2 graphics looks something like this:
ggplot(data=, aes(x=,y=, [options]))+geom_xxxx()+...+...+...
Note that ggplot2 graphs in layers in a continuing call (hence the endless +…+…+…), which really makes the extra layer part of the call
...+geom_xxxx(data=, aes(x=,y=,[options]),[options])+...+...+...
You can see the layering effect by comparing the same graph with different colors for each layer
ggplot(data=data, aes(x=year, y=realgdpgr))+geom_point(color="black")+geom_point(aes(x=year, y=unemp), color="red")
ggplot(data=data, aes(x=year, y=realgdpgr))+geom_point(color="red")+geom_point(aes(x=year, y=unemp), color="black")
densityplot(~vturn, data=data) # lattice
ggplot(data=data, aes(x=vturn))+geom_density() # ggplot2
xyplot(outlays~year, data=data) # lattice
ggplot(data=data, aes(x=year, y=outlays))+geom_point() # ggplot2
xyplot(outlays~year, data=data[data$country=="USA",], type="l") # lattice
ggplot(data=data[data$country=="USA",], aes(x=year, y=outlays))+geom_line() # ggplot2
# Create data.frame of average growth rates by country over time
growth<-ddply(.data=data, .variables=.(country), summarize, mean=mean(realgdpgr, na.rm=T))
barchart(mean~country, data=growth) # lattice
ggplot(data=growth, aes(x=country, y=mean))+geom_bar() # ggplot2
bwplot(outlays~country, data=data) # lattice
ggplot(data=data, aes(x=country, y=outlays))+geom_boxplot() # ggplot2
xyplot(outlays~year|country, data=data) # lattice
ggplot(data=data, aes(x=year, y=outlays))+geom_point()+facet_wrap(~country) # ggplot2
data(volcano) # Load volcano contour data
volcano[1:10, 1:10] # Examine volcano dataset (first 10 rows and columns)
## [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10]
## [1,] 100 100 101 101 101 101 101 100 100 100
## [2,] 101 101 102 102 102 102 102 101 101 101
## [3,] 102 102 103 103 103 103 103 102 102 102
## [4,] 103 103 104 104 104 104 104 103 103 103
## [5,] 104 104 105 105 105 105 105 104 104 103
## [6,] 105 105 105 106 106 106 106 105 105 104
## [7,] 105 106 106 107 107 107 107 106 106 105
## [8,] 106 107 107 108 108 108 108 107 107 106
## [9,] 107 108 108 109 109 109 109 108 108 107
## [10,] 108 109 109 110 110 110 110 109 109 108
volcano3d <- melt(volcano) # Use reshape2 package to melt the data
head(volcano3d) # Examine volcano3d dataset (head)
## Var1 Var2 value
## 1 1 1 100
## 2 2 1 101
## 3 3 1 102
## 4 4 1 103
## 5 5 1 104
## 6 6 1 105
names(volcano3d) <- c("xvar", "yvar", "zvar") # Rename volcano3d columns
contourplot(zvar~xvar+yvar, data=volcano3d) # lattice
ggplot(data=volcano3d, aes(x=xvar, y=yvar, z = zvar))+geom_contour() # ggplot2
levelplot(zvar~xvar+yvar, data=volcano3d) # lattice
ggplot(data=volcano3d, aes(x=xvar, y=yvar, z = zvar))+geom_tile(aes(fill=zvar)) # ggplot2
# Create a subset of the dataset containing only data for France
france.data<-data[data$country=="France",]
cloud(outlays~year*realgdpgr, data=france.data)
# Create a subset of the dataset containing only data for Greece, Portugal, Ireland, and Spain
pigs.data<-data[data$country %in% c("Greece", "Portugal", "Ireland", "Spain"),]
cloud(outlays~year*realgdpgr|country, data=pigs.data)
ggplot(data=pigs.data, aes(x=year, y=realgdpgr, color=country))+geom_line()
xyplot(outlays~year, data=data, xlab="Year", ylab="Government Outlays", main="Cool Graph") # lattice
ggplot(data=data, aes(x=year, y=outlays))+geom_point()+xlab(label="Voter Turnout (%)")+ylab(label="Government Outlays")+ggtitle(label="Cool Graph") # ggplot2
xyplot(outlays~year, data=data) # lattice
xyplot(outlays~year, data=data, cex=2) # lattice
xyplot(outlays~year, data=data, cex=.5) # lattice
ggplot(data=data, aes(x=year, y=outlays))+geom_point() # ggplot2
ggplot(data=data, aes(x=year, y=outlays))+geom_point(size=3) # ggplot2
ggplot(data=data, aes(x=year, y=outlays))+geom_point(size=1) # ggplot2
xyplot(outlays~year, data=data, col=colors()[145]) #lattice
xyplot(outlays~year, data=data, col="red") #lattice
ggplot(data=data, aes(x=year, y=outlays))+geom_point(color=colors()[145]) # ggplot2
ggplot(data=data, aes(x=year, y=outlays))+geom_point(color="red") # ggplot2
xyplot(outlays~year, data=data, pch=3) # lattice
xyplot(outlays~year, data=data, pch=15) # lattice
ggplot(data=data, aes(x=year, y=outlays))+geom_point(shape=3) # ggplot2
ggplot(data=data, aes(x=year, y=outlays))+geom_point(shape=15) # ggplot2
xyplot(outlays~year, data=data, pch=3) # lattice
xyplot(outlays~year, data=data, pch=15) # lattice
xyplot(outlays~year, data=data, pch="w") # lattice
xyplot(outlays~year, data=data, pch="$", cex=2) # lattice
ggplot(data=data, aes(x=year, y=outlays))+geom_point(shape=3) # ggplot2
ggplot(data=data, aes(x=year, y=outlays))+geom_point(shape=15) # ggplot2
ggplot(data=data, aes(x=year, y=outlays))+geom_point(shape="w") # ggplot2
ggplot(data=data, aes(x=year, y=outlays))+geom_point(shape="$", size=5) # ggplot2
xyplot(outlays~year, data=data[data$country=="USA",], type="l", lty=1) # lattice
xyplot(outlays~year, data=data[data$country=="USA",], type="l", lty=2) # lattice
xyplot(outlays~year, data=data[data$country=="USA",], type="l", lty=3) # lattice
xyplot(outlays~year, data=data[data$country=="USA",], type="l", lty=3, lwd=2) # lattice
xyplot(outlays~year, data=data[data$country=="USA",], type="l", lty=3, lwd=3) # lattice
xyplot(outlays~year, data=data[data$country=="USA",], type="l", lty=3, lwd=4) # lattice
ggplot(data=data[data$country=="USA",], aes(x=year, y=outlays))+geom_line(linetype=1) # ggplot2
ggplot(data=data[data$country=="USA",], aes(x=year, y=outlays))+geom_line(linetype=2) # ggplot2
ggplot(data=data[data$country=="USA",], aes(x=year, y=outlays))+geom_line(linetype=3) # ggplot2
ggplot(data=data[data$country=="USA",], aes(x=year, y=outlays))+geom_line(linetype=3, size=1) # ggplot2
ggplot(data=data[data$country=="USA",], aes(x=year, y=outlays))+geom_line(linetype=3, size=1.5) # ggplot2
ggplot(data=data[data$country=="USA",], aes(x=year, y=outlays))+geom_line(linetype=3, size=2) # ggplot2
By now, you might be noticing some trends in how these two packages approach graphics
lattice tends to focus on a particular type of graph and how to represent cross-sectional variation by splitting it up into smaller chunks
Becoming a proficient user of lattice requires learning a huge array of graph-specific formulas and options
ggplot2 tries to represent much more of the cross-sectional variation by making use of various “aesthetics”; general approach is based on The Grammar of Graphics
1) One or more statistics conveying information about the data (identities, means, medians, etc.)
2) A coordinate system that differentiates between the intersections of statistics (at most two for ggplot, three for lattice)
3) Geometries that differentiate between off-coordinate variation in kind
4) Scales that differentiate between off-coordinate variation in degree
ggplot(data=, aes(x=, y=, color=, linetype=, shape=, size=))
ggplot2 is optimized for showing variation on all four aesthetic types
# Differences in kind using color
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_line(aes(color=country))
Note what happens when we specify the color parameter outside of the aesthetic operator. ggplot2 views these specifications as invalid graphical parameters.
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_line(color=country)
## Error: object 'country' not found
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_line(color="country")
## Error: invalid color name 'country'
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_line(color="red")
# Differences in kind using line types
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_line(aes(linetype=country))
# Differences in kind using point shapes
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point(aes(shape=country))
# Differences in degree using color
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point(aes(color=realgdpgr))
# Differences in degree using point size
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point(aes(size=realgdpgr))
# Multiple non-cartesian aesthetics (differences in kind using color, degree using point size)
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point(aes(color=country,size=realgdpgr))
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point()
# Add linear model (lm) smoother
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point()+geom_smooth(method="lm")
# Add local linear model (loess) smoother, span of 0.75
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point()+geom_smooth(method="loess", span=.75)
# Add local linear model (loess) smoother, span of 0.25
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point()+geom_smooth(method="loess", span=.25)
# Add linear model (lm) smoother, no standard error shading
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point()+geom_smooth(method="lm", se=F)
# Add local linear model (loess) smoother, no standard error shading
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point()+geom_smooth(method="loess", se=F)
# Add a local linear (loess) smoother for each country
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point(aes(color=country))+geom_smooth(aes(color=country))
# Add a local linear (loess) smoother for each country, no standard error shading
ggplot(data=pigs.data, aes(x=year, y=outlays))+geom_point(aes(color=country, size=realgdpgr))+geom_smooth(aes(color=country), se=F)
# Initialize gridExtra library
library(gridExtra)
# Create 3 plots to combine in a table
plot1<-ggplot(data=pigs.data, aes(x=year, y=outlays, color=))+geom_line(aes(color=country))
plot2<-ggplot(data=pigs.data, aes(x=year, y=outlays, linetype=))+geom_line(aes(linetype=country))
plot3<-ggplot(data=pigs.data, aes(x=year, y=outlays, shape=))+geom_point(aes(shape=country))
# Call grid.arrange
grid.arrange(plot1, plot2, plot3, nrow=3, ncol=1)
Two basic image types
1) Raster/Bitmap (.png, .jpeg)
Every pixel of a plot contains its own separate coding; not so great if you want to resize the image
jpeg(filename="example.png", width=, height=)
plot(x,y)
dev.off()
2) Vector (.pdf, .ps)
Every element of a plot is encoded with a function that gives its coding conditional on several factors; great for resizing
pdf(filename="example.pdf", width=, height=)
plot(x,y)
dev.off()
# Assume we saved our plot is an object called example.plot
# lattice
trellis.device(device="pdf", filename="example.pdf")
print(example.plot)
dev.off()
# ggplot2
ggsave(filename="example.pdf", plot=example.plot, scale=, width=, height=) # ggplot2
1) Not all variable types are suitable for representation by every ggplot aesthetic. What kinds of variables can the aesthetics color, size, and shape meaningfully represent?
2) Using ggplot2, create a trellis plot where, for a given country, each panel uses a) HOLLOW CIRCLES to plot real GDP growth over time, and b) a red LOESS smoother without standard errors to plot the trend in unemployment over time. BONUS: limit the years shown to the period from 2000 to 2010, and turn off the grey background.